Академический Документы
Профессиональный Документы
Культура Документы
3390/ma7064854
OPEN ACCESS
materials
ISSN 1996-1944
www.mdpi.com/journal/materials
Article
Abstract: In this study, the scope of CFRP cables in cable-stayed bridges is studied by
establishing a numerical model of a 1400-m span of the same. The mechanical properties
and characteristics of CFRP stay cables and of a cable-stayed bridge with CFRP cables are
here subjected to comprehensive analysis. The anomalies in the damping properties of free
vibration, nonlinear parametric vibration and wind fluctuating vibration between steel
cables and CFRP cables are determined. The structural stiffness, wind resistance and traffic
vibration of the cable-stayed bridge with CFRP cables are also analyzed. It was found that
the static performances of a cable-stayed bridge with CFRP cables and steel cables are
basically the same. The natural frequencies of CFRP cables do not coincide with the major
natural frequencies of the cable-stayed bridge, so the likelihood of CFRP cable-bridge
coupling vibration is minuscule. For CFRP cables, the response amplitudes of both
parametric vibration and wind fluctuating vibration are smaller than those of steel cables. It
can be concluded from the research that the use of CFRP cables does not change the
dynamic characteristics of the vehicle-bridge coupling vibration. Therefore, they can be
used in long-span cable-stayed bridges with an excellent mechanical performance.
Keywords: CFRP stay cables; long span; cable-stayed bridge; static characteristics;
dynamic characteristics
Materials 2014, 7
4855
1. Introduction
Cable-supported bridges have undergone rapid development over the last century since high-strength
steel cables were first made available. Cable-stayed bridges can reach spans of over 1000 m, which makes
them practical alternatives to suspension bridges. The cables, which transmit most of the force, are key
to the structural stability and safety of cable-stayed bridges. However, cables are very vulnerable to
corrosion and fatigue damage, which affect their service lifespan, necessitating their replacement. In
China alone, the cables of more than twenty bridges have been replaced over the past twenty years for
safety reasons [1]. Cable replacement not only requires a huge financial investment, but also disrupts
transportation and causes problems in the maintenance and management of the bridge.
Carbon fiber-reinforced plastic (CFRP) has been widely used in civil engineering, because of its
light weight, great strength, considerable flexibility and resistance to corrosion and fatigue. In spite of
the disadvantages, like the lack of ductility, high cost, contractor unfamiliarity, susceptibility to impact
damage and difficulty in forming connections, CFRP cables are superior to steel cables in some critical
mechanical aspects, such as creep and relaxation. The possibility of using CFRP in long-span bridges
has received considerable attention for this reason. Meier et al. [2] evaluated the applicability of CFRP
materials in long-span cable-supported bridges. Later, other scholars made more comprehensive studies
on the possibility of applying CFRP in long-span bridges [314]. The mechanical properties of
long-span cable-supported bridges built with CFRP cables have been gradually explained. Better ways
of manufacturing CFRP cables and anchoring systems have been developed, and these have already
seen use in some real-world bridges [1519]. The increase in knowledge of CFRP materials has
decreased the manufacturing costs, and the use of CFRP cables in long-span bridges will upsurge in the
near future.
So far, most studies of CFRP cables have been limited in scientific premise. Some problems are still
at the probing stage. The impact of CFRP cables on the mechanical properties of long-span
cable-stayed bridges has not yet been subjected to systematic research. In this research, a series of
studies have been carried out on CFRP cables in bridges of different sizes. Herein, a CFRP
cable-stayed bridge is compared to a steel cable-stayed bridge with respect to material, element and
structural characteristics. These results provide evidence supporting the use of CFRP cables in
long-span cable-stayed bridges.
2. Mechanical Properties of CFRP Cables
2.1. Material Properties of CFRP Cables
At present, there is no uniform production standard for CFRP cables, so the material parameters of
CFRP cables produced by different manufacturers differ considerably. Table 1 shows the material
properties of four types of CFRP cables, marked as A, B, C and D. In the table, , E, u, u, Re and
represent the unit weight, Youngs modulus, tensile strength, ultimate strain, relaxation ratio for 1000 h
and linear expansion coefficient, respectively. It was here observed that the material characteristics
of CFRP cables produced by different manufacturers are typically similar. The Youngs modulus is
about 140 GPa, which is about 70% of steel cables. The tensile strength is 2.022.55 GPa, which
is 1.31.6 times that of high-strength steel cables. The linear expansion coefficient is 0.6 106/C, and
Materials 2014, 7
4856
the temperature deformation is only 1/20 that of steel cables. The unit weight is 1/5 that of steel cables,
and the relaxation rate is lower than that of steel cables. In addition, the ultimate strain of CFRP cable is
so small that the stress-strain curve remains mostly linear before fissure, which indicates that CFRP
cables are brittle.
Table 1. Material properties of CFRP and steel cables.
Cable types
A
B
C
D
steel
(kN/mm3)
16
16
16
16
77
u (MPa)
2,140
2,550
2,022
2,421
1,570
E (GPa)
137
147
137
159
196
u /E
0.016
0.017
0.015
0.015
0.008
Re (%)
0.3
0.3
0.3
0.3
<2.5
u (%)
1.6
1.6
2
1.7
>4
al (106/C)
0.6
0.68
0.6
0.6
12
1 W
4 W
(1)
Here, W is the energy absorbed due to the strain deformation during the vibration cycle, and W is the
modal potential energy. Assuming W is proportional to the strain energy WV, replacing the
proportionality sign with a constant [2022]:
W 2V
(2)
Here, is the energy loss factor, and V is the strain energy. Substitution of the equation above into
Equation (1) produces the following:
1 V
2 W
(3)
Because there is a large initial tension T0 in the cable, therefore the modal potential energy W of
Formula (3) includes two parts, initial tension potential energy W0 and strain energy V, hence:
W W0 V
(4)
Because the flexible cable only bears axial tension, as indicated by the relationship of strain and
tension shown in Figure 1, the initial potential energy due to tension, W0 and strain energy V can be
calculated as follows:
n
W0i Tk 0 S k ki ,
k 1
n
EASk 2
Vi
ki .
2
k 1
(5)
Materials 2014, 7
4857
Figure 1. Potential energy and strain energy in initial tension.
T
T
T0
W0
Here, ki and Sk are the strain and length of element k generated by the i-th vibration mode. Tk0 is the
initial tension in element k, and n is the total number of cable elements. Substitution of Equations (4)
and (5) into (1) results in the following expression:
i
1
Vi
2 W0i Vi
(6)
In order to compare the damping characteristics of steel and CFRP cables, a cable vibration test was
carried out in Tokyo Metropolitan University [22]. The test steel and CFRP cables were made at Rope
Corporation Tokyo, Japan, wherein the steel cable, PCS12.7 ( refers to diameter of the cable), was
composed of seven 5.16-mm diameter steel wires, PCS5.16, and the CFRP cable, CFCC12.5, was
composed of seven 5.0 mm-diameter CFRP wires, CFCC5.0 (Figure 2). Table 2 lists the material
parameters of cables and wires. E, w, D, A and u represent Youngs modulus, the unit weight, the
diameter, the cross-sectional area and the tensile strength of cables and wires, respectively.
Figure 2. Cross-section of cable CFCC12.5.
cable
wires
E (GPa)
159
197
152
204
w (g/m)
145
774
29.9
163.8
D (mm)
12.50
12.70
5.00
5.16
A (mm2)
76.0
98.7
15.2
20.9
u (MPa)
2,421
1,854
2,980
1,999
Figure 3a,b provides details of the cable damping test equipment. The test cable was 20 m-long
with an initial tension of 30 kN. The resulting initial stresses in CFRP cables and steel cables were 395
and 304 MPa, respectively, which were equal to half of the service stresses. Here, the design safety
Materials 2014, 7
4858
factor of the test cable was 2.5. Continuous excitation force was applied at the midpoint of the cable to
render the vibrations stable. Once the excitation force ceased, the cable entered free decay vibration,
which was measured by a non-contact laser displacement meter at a sampling rate of 500 Hz. Figure
3b shows the relationship between the cable damping and vibration amplitude. It can be found that, for
steel and CFRP cables, damping increases with vibration amplitude, but the trend was more obvious
for steel cables than CFRP cables. This indicted that the influence of vibration amplitude on cable
damping is not as pronounced for CFRP cable as it is for the steel cable. In addition, the energy loss
factor of cables shown in Equation (3) can be calculated from Figure 3b. Here, the damping constant
was 0.05 for both steel and CFRP cables, consistent with the test results reported by Kady et al. [23].
Figure 3. (a) Cable damping test; (b) cable damping log decrement vs. vibration amplitude.
0.006
Test cable
Sag
T
laser displacement meter
Tension
equipment
Sine wave
vibration exciter
(a)
(b)
CFRP-analytical
Steel-analytical
CFRP-experimental
0.004
Steel-experimental
0.002
0.000
0
1
2
3
vibration amplitude d max (mm)
In order to determine the energy loss factors of wires comprising the cables, the bending
deformation experiment of steel and CFRP wires under free vibration was conducted (Figure 4a). A
heavy block was hung on by a string on the fixed test wire at a distance of l from the fixed end. An
initial imposed displacement was used to place the wire in decay free vibration. The strain readings
were recorded by the dynamic strain gauges fixed on the wire, at a sampling of 500 Hz. During the
test, the distance l and the mass of the heavy block were varied multiple times to obtain additional
experimental data.
Figure 4. (a) Bending deformation experiments of steel and CFRP wires; (b) distribution of
energy loss coefficient for steel wires; (c) distribution of energy loss coefficient for CFRP
wires.
80
(a)
y
l
60
(c)
60
Steel wire
O
String
CFRP wire
Count
Count
Count
Count
CFRP (steel)
wires
CFRP
Strain gauges
Fixture
80
(b)
40
20
Heavy block
0
0.01
40
20
0.02
0.03
0
0.00
0.01
0.02
Figure 4b shows the distribution results of energy loss factors for steel and CFRP wires. As shown,
the average energy loss factors for steel and CFRP wires are about 0.018 and 0.008, respectively,
which is much smaller than the damping constant of steel and CFRP cables, 0.05. This indicates that
Materials 2014, 7
4859
the damping of the material strain ratio is only a part of the total cable damping, and other damping is
relative to the friction among the wires of the cable.
3. Structural Description of the Cable-Stayed Bridge
In this paper, the mechanical characteristics of a cable-stayed bridge with CFRP cables are discussed,
and the feasibility of using CFRP cables on long-span cable-stayed bridges is evaluated.
The finite element method (FEM) is used to establish a numerical model of the 1400-m span of a
cable-stayed bridge. Although the bridge model is not an actual structure, the geometric parameters
were chosen by a conceptual design procedure according to some basic parameters of the existing
long-span cable-stayed bridges [22]. Figure 5 shows the layout of the bridge model (in meters). This
bridges mid-span is 1400 m, which is about two times its side-span. Thirty-four cables are installed at a
horizontal spacing of 20 m on each side of towers to hold the main beam, which is a uniform multi-cell
flat steel box girder. The width of the box girder is 30 m, and the height is 4.5 m, which indicates a
relatively larger height-to-span ratio as compared to many other cable-stayed bridges. The equivalent
thicknesses of the orthotropic top and bottom plates of the girder are 20 mm (derived as
12 mm of steel plate and an 8-mm equivalent thickness of the stiffener). The equivalent thickness of the
web is 15 mm. The height of the pylon is 280 m, which is 0.2-times the mid-span, and the tower
cross-section is a single cell box composed of walls with an equivalent thickness of 40 mm. Table 3 lists
the cross-sectional parameters of the girder and the tower. Along the side-span, there are three middle
auxiliary piers at a horizontal spacing of 100 m to support the girder. Constraint springs were set
between the girder and the pylon along the bridge longitudinal axis. Cables of C1, C2 and C3, shown in
Figure 5, represent long, medium and short cables of Type C, mentioned in Table 1. The cables here
have a relatively low tensile strength of 2022 MPa and an elastic modulus of 137 kN/mm2. In the
numerical model, the tower and box girder are 1111 spatial beam elements with 6 DOF, and the cables
are 272 isoparametric elements with 3 DOF. The bridge is fixed at the bottom of the tower and supported
by piers. The bridge can move along the longitudinal axis. The geometric nonlinear method and material
nonlinear method are used to solve the problem.
The unit weight of the structure is calculated as follows:
WD kD A W0
(7)
Here, kD is the load factor, chosen by conceptual design considering the self-weight of the stiffener
plate of the box girder. [22]. For girders kD = 1.4, and for pylons, kD = 1.2. is the unit weight of steel.
A is the cross-sectional area of box girder. W0 is the secondary load, which mainly consists of
pavement load, etc., taken as, W0 = 70 kN/m. The secondary load of the pylon is relatively so small,
that it can be ignored.
Table 3. Cross-sectional properties.
Cross-Section
A (m2)
Ix (m4)
Iy (m4)
Girder
Tower
1.607
1.760
5.503
30.667
120.43
40.320
10.352 *
39.273 *
Materials 2014, 7
4860
Figure 5. (a) 3D model of the bridge; (b) layout of half of the bridge; (c) typical
cross-section of the box girder; (d) layout of the pylon and cross-section of the tower.
(a)
(b)
C3
100 100
100
18@20=360
1820=360
20
C2
20
20
33@20=660
33
20=660
700
680
(c)
C1
30
2%
2%
4.5
10
148
10
280
33
4=132
33@4=132
(d)
47.5
40
12
61
The cross-sectional area of the cable is calculated using the maximum cable tension T0 and the
allowable stress . The formula is as follows:
Ac
1 T0
1 T0
b
(8)
Materials 2014, 7
4861
Here, is the factor of proportionality. It indicates how much cable tension is produced by the live
load in response to the maximum tension produced by the dead load in the long-span cable-stayed
bridge. Because the factor of proportionality for long-span cable-stayed bridges is usually not more
than 0.2, it is here assumed that = 0.2. is the allowable cable stress, and b is the standard value of
tensile strength; is the safety factor for the cable, which is generally 2.5 for steel cables. It is here assumed
that = 3.0 for CFRP cables, because CFRP is a strong, but brittle, material.
Table 4 shows the cable self-weight for steel cable-stayed bridges and CFRP cable-stayed bridges
of the same span. Results show that the self-weight of the CFRP cable is only about 1/7 that of steel
cable, which is less than the unit weight ratio of CFRP to steel, which is about 1/5. This means that
fewer CFRP than steel cables would be needed for a cable-stayed bridge. This is because CFRP cables
sag less, giving them greater working efficiency. This may substantiate the cost disadvantage of
CFRP cables.
Table 4. Weight of steel and CFRP cables on a whole bridge.
Material
Weight (kN)
Steel
CFRP
115,133
17,056.7
6.8
wL
H
LEA
HLe
(9)
Here, H represents the horizontal component of cable tension and Le represents cable length.
Expression (9) indicates that the Irvine parameter is related to the geometric and material properties of
the cable, the smaller the sag and the tensile stiffness of the cable, the smaller the Irvine parameter.
Table 5 shows that, for the same span cables, the Irvine parameters of steel cables are much larger than
those of CFRP cables. For cables made of the same material, the Irvine parameters for long cables are
larger than that for short cables. Table 6 shows that, for the same cable span, the natural frequencies of
CFRP cables are about two-times those of steel cables.
Materials 2014, 7
4862
Figure 6. Natural frequencies of steel and CFRP cables.
5
f (Hz)
Steel-1st
4
Steel-2nd
CFRP-1st
CFRP-2nd
2
1
0
0
10
20
30
40
50
60
70
Cable numbers
L (m)
660
460
260
660
460
260
Y (m)
275.5
235.5
195.5
275.5
235.5
195.5
EA (kN)
3,045,840
2,516,640
1,795,360
1,557,416
1,377,507
1,033,802
W (kN/m)
1.3114
1.0835
0.7730
0.1840
0.1630
0.1220
0.01750
0.01306
0.00912
0.002863
0.002144
0.001501
Out-of-Plane
1st
2nd
0.1652
0.3343
0.2289
0.4635
0.3645
0.7379
0.4083
0.8267
0.5662
1.1443
0.8986
1.8194
In-Plane
1st
2nd
0.1998
0.3341
0.2532
0.4633
0.3771
0.7378
0.4098
0.8266
0.5652
1.1443
0.8991
1.8194
In order to analyze the relationship between the cables and the dynamics of the bridge itself with
respect to natural frequencies, the major frequencies of the bridge were calculated taking the influence
of local vibrations of the cable into account (Table 7). The results show that the natural frequencies of
steel cable-stayed bridges are basically the same as those of CFRP cable-stayed bridges.
Table 7. Major natural frequencies of cable-stayed bridge (units: Hz).
Cable Type
Steel cables
CFRP cables
Longitudinal Drifting
0.0330
0.0340
It can be found from Table 6 and Table 7 that the natural frequencies of steel cables overlap those
of the first vertical bending frequencies of steel cable-stayed bridge. However, the natural frequencies
Materials 2014, 7
4863
of CFRP cables differ from the low natural frequencies of the CFRP cable-stayed bridge. That is why
the steel cable-bridge coupling vibrations occur so easily. Therefore, using CFRP cables can
significantly reduce the possibility of coupling vibrations in the cable bridge. Some major vibration
modes of steel cable-stayed bridges and CFRP cable-stayed bridges are shown in Figure 7. It was
found that the local vibrations of the cables in steel cable-stayed bridges were dominant for the first
vertical bending vibration.
Figure 7. (a) Vibration modes of cable-stayed bridges with steel cables; (b) vibration
modes of cable-stayed bridges with CFRP cables.
(a)
(b)
Materials 2014, 7
4864
Figure 7. Cont.
(b)
4.2. Damping Characteristics
Because the strain energy ratios of different lengths of cables are similar, only cable C2 is analyzed
in this paper. Figure 8 shows the relationship between the first two vibration modal strain energy ratios
calculated using Formulas (5) and (6) and the cable lateral in-plane and out-of-plane amplitudes. In the
figure, S represents the steel cable, C represents the CFRP cable, -1st and -2nd represent the first and
the second vibration modes and -out and -in represent the in-plane and out-of-plane vibration modes.
The results show that the modal strain energy ratios of two materials are on the same order of
magnitude. CFRP cables showed lower strain energy ratios than steel cables. If the vibration amplitude
is around 1 m, Formula (6) can be used to calculate the structural damping ratios of both steel and
CFRP cables, which were here around 1%.
Figure 8. Strain energy ratios of steel and CFRP cables.
0.3
C-1st-out
C-2nd-out
S-1st-out
S-2nd-out
0.2
C-1st-in
C-2nd-in
S-1st-in
S-2nd-in
0.1
0.0
0
The external viscous damper is usually arranged at the lower end of the cable in order to prevent
large vibration. Figure 9 shows the layout of the cable with the external viscous damper. xc is the
Materials 2014, 7
4865
horizontal distance between the lower end of the cable and the damper. In this paper, xc = 0.02 L. The
non-proportional damping calculation theory (complex eigenvalue method) was used to calculate the
damping value.
Figure 9. Stay cable with viscous damper.
Cross-section of cable
C
Viscous damper
C
xc=0.02L
The calculation results show that the damping coefficients C have little impact on the natural
frequencies of the steel and CFRP cables. Figure 10 shows the relationship between the first in-plane
modal damping ratios and the damping coefficients. The damping coefficients significantly affected
the modal damping ratios. As the damping coefficients increase, the modal damping ratios also
increase significantly, peak and then drop quickly. There is an optimal range within which the
damping coefficients produce desirable effects. The optimal damping coefficient can be found using
the following formula [25].
C 0.1
m1 L2
xc
(10)
Here, m represents the unit mass and 1 represents the in-plane damped natural circular frequency of
the first mode of the stay cable. Comparisons showed that the optimal viscous damping coefficients of
CFRP cables were smaller than that of steel cables. This shows that only smaller dampers can produce
the similar effects on CFRP cables that have been seen on steel cables.
Figure 10. Effects of external dampers on modal damping.
0.010
0.008
S-C1
S-C2
S-C3
C-C1
C-C2
C-C3
0.006
0.004
0.002
0.000
0
200
400
600
800
C (kN.s/m)
1000
1200
1400
Materials 2014, 7
4866
Sag
Steel
f1
28.460
11.590
4.007
CFRP
2 f1
10.670
11.150
3.680
f1
7.059
4.270
1.961
2 f1
7.410
4.946
1.961
Materials 2014, 7
4867
30
20
10
0
-10
-20
-30
(a)
dy(m)
S-C1-f1
100
200
300
400
500
600
30
20
10
0
-10
-20
-30
(b)
700
S-C1-2f1
100
200
300
10
10
500
600
700
C-C1-2f1
(d)
dy(m)
dy(m)
C-C1-f1
(c)
400
t (s)
t (s)
0
-5
-5
-10
-10
0
100
200
300
400
500
100
200
300
400
500
t (s)
t (s)
20
(a)
S-C1-f1-damped
(b)
S-C1-2f1-damped
10
dy(m)
dy(m)
10
0
-10
0
-10
-20
0
100
200
300
400
500
600
700
-20
0
100
200
300
t (s)
400
500
600
700
t (s)
10
10
(c)
(d)
C-C1-f1-damped
C-C1-2f1-damped
dy(m)
5
dy(m)
(d)
dy(m)
Figure 12. Parametric vibration response of stay cable C1 (undamped): (a) steel cable
under excitation frequency f = f1; (b) steel cable under excitation frequency f = 2f1;
(c) CFRP cable under excitation frequency f = f1; (d) CFRP cable under excitation
frequency f = 2f1.
0
-5
0
-5
-10
0
100
200
300
t (s)
400
500
-10
0
100
200
300
t (s)
400
500
Materials 2014, 7
4868
4.4. Fluctuating Vibration Characteristics of CFRP Stay Cables under Wind Load
4.4.1. Calculation
The analysis of wind loads to which the cables are subjected is very complex. The fluctuating wind
load can cause them to vibrate. The turbulence power spectral density function of fluctuating wind is
as follows [37]:
Su f
4 I u2U 2 Lx U
1 70.8 f L
(11)
5
2 6
Lx 100Z 30
0.5
(12)
Here, Su f represents the turbulence power spectral density function; Lx represents the scale of
turbulence; Iu represents turbulence intensity, which is related to wind speed, and the ground roughness
coefficient Iu = 0.1 in this paper. Z is the distance of the girder from the surface of the water surface.
Here, Z = 50 m.
The random velocity of fluctuating wind is as follows [38]:
n
U t ak cos k t k
(13)
k 1
Here, k is the randomly-generated phase angle, and other parameters are calculated as follows:
fu fl
f n
ak 2 S f k f
f f k 1 f
l
k
2
k 2f k
k 1, 2,
, n
(14)
Here, fu and fl are the upper and lower bounds of cable frequencies and n is an integer. The power spectral
density of fluctuating wind sample with an average speed of 30 m/s was generated using Equation (11).
For the cable laden by lateral wind load shown in Figure 14, the relative wind attack angle can be
determined as follows:
tan
Uy y
Uz z
(15)
Here, Uy and Uz are vertical and horizontal components of fluctuating wind speed, which are calculated
as follows:
U y U sin
U y U cos
(16)
Here, is the wind attack angle shown in Figure 14; y and z are the vertical and horizontal
vibration components, respectively. U is the sum of average wind speed and fluctuating wind speed.
Materials 2014, 7
4869
Figure 14. Cross-section of cable under lateral wind load.
z
U
Uz
y U z z
2
(17)
The drag force D and lift force L acting on the cable along the relative wind direction are as follows:
1
2
D 2 CD U r
L 1 C U 2
L
r
(18)
Here, CD and CL are the drag and lift coefficients of the cable. CD = 0.7 and CL = 0 for the round
cross-section of the cable. is the outside diameter of the cable. is the air density. The vertical and
horizontal wind loads can be compounded by drag and lift force:
qz D cos L sin
q y D sin L cos
(19)
(a)
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
(b)
C-C1
dy(m)
dy(m)
S-C1
50
100
150
t (s)
200
250
50
100
150
t (s)
200
250
Materials 2014, 7
4870
Figure 15. Cont.
0.2
0.2
(c)
(d)
S-C2
dy(m)
dy(m)
C-C2
0.1
0.1
0.0
-0.1
0.0
-0.1
-0.2
0
50
100
150
200
250
-0.2
0
50
100
t (s)
200
250
200
250
0.1
0.1
(e)
(f)
S-C3
dy(m)
dy(m)
150
t (s)
0.0
C-C3
0.0
-0.1
-0.1
0
50
100
150
200
250
t (s)
50
100
150
t (s)
(20)
Here, u is the displacement vector; K(u) is the stiffness matrix considering the initial internal force and
geometrical nonlinearity; and F is the live load vector. To determine the most unfavorable loading
position of the live load, the tangent stiffness at the finished state of the bridge is used to identify the
influence line. The reduction of the live load with the span is neglected.
Figure 16 gives the maximum and minimum deflection values of a half box girder. The minimum
deflection of the CFRP cable-stayed bridge was nearly same as that of the steel cable-stayed bridge.
However, because steel-cable-stayed bridges have greater structural rigidity, the maximum girder
deflection is 0.000119 of the span, but for CFRP cable-stayed bridges, it is 0.000146 of the span, which
is 1.22-times the former. This indicates that the material properties of the cable have little impact on the
minimum girder deflection and considerable impact on the maximum deflection. The critical girder
deflection of cable-stayed bridge is 0.001 of the span, meaning that replacing steel cables with CFRP
cables might give the bridge a large margin of safety, with respect to structural rigidity.
Under dead and live loads, the maximum compressive stress in the girders of steel-cable-stayed
bridges is 189.3 MPa; that of CFRP-cable-stayed bridges it is 9% lower, at 173.9 MPa. This means that
CFRP cables may be suitable for long-span cable-stayed bridges.
Structural stability is very important for long-span cable-stayed bridges, because there is a large
amount of axial pressure in girders and pylons. The stability safety factor can be obtained using the
following complex eigenvalue equation [40]:
det K ep K G 0
(21)
Materials 2014, 7
4871
Here, Kep is the stiffness matrix, which takes material nonlinearity into consideration; KG is the
geometric stiffness matrix; and is the stability safety factor. Unlike the linear elastic eigenvalue
method, the elastic modulus in Equation (21) was obtained based on the law that the buckling strength is
equal to the experimental strength, i.e., the effective elastic modulus was adopted for the stiffness matrix,
which considers the influence of initial defect and elasto-plastic stress of the structural member.
Figure 16. Maximum and minimum girder deflections under live load.
0.25
Deflection Dx (m)
CFRP Dx,max
0.20
CFRP Dx,min
0.15
STEEL Dx,max
0.10
STEEL Dx,min
0.05
0.00
-0.05
0
300
600
x (m)
900
1200
1500
Elasto-plastic branch instability analysis showed that the minimum stability factor is 3.03 for CFRP
cable-stayed bridges and 2.88 for steel cable-stayed bridges. The difference is attributed to the buckling
caused primarily by the pylon. The self-weight of the steel cables is greater than that of CFRP cables, so
the axial force in the pylon for steel-cable-stayed bridge is greater than that for a CFRP-cable-stayed
bridge. Figure 17 gives an example of the elasto-plastic buckling mode for cable-stayed bridge with
CFRP cables.
Figure 17. Elasto-plastic buckling mode of cable-stayed bridge.
2
L 0.5U BCL
2 2
M 0.5U B CM
(22)
Materials 2014, 7
4872
Here, is the wind attack angle; is air density; B is girder width; An is girder height in the windward
side; and U is wind speed. Here, U = 40 m/s. D, L and M are the drag force, the lift force and the torque
action on unit length of the box girder in relation to the drag coefficient CD, the lift coefficient CL and
the torque coefficient CM, respectively. Figure 18 gives the relation curves between the three aerostatic
coefficients and the wind attack angles.
CD (CL ,CM )
Figure 18. Correlation between aerostatic coefficients and the wind attack angles.
4
3
CD
2
CL
CM
0
-15
-10
-5
-1
10
Deg
15
-2
Figure 19 gives the transverse bending moment and deformation of the half-bridge girder under the
design wind load. Due to the constraints imposed by the middle piers, the bending moment and
deformation of the side span girders are very small, but that of the mid-span girder is considerably high.
The transverse bending moment and deformation of mid-span girders is slightly larger for steel
cable-stayed bridges than that for CFRP cable-stayed bridges. The logic behind this is that the
out-of-plane deformation of the girder is related to the windward area of the cable, and steel cables have
more windward area than CFRP cables.
The analysis of the instability of cable-stayed bridges under static wind load showed that when the
wind speed reaches 60 m/s for a complete bridge or 50 m/s for a bridge still under construction, bridges
built with CFRP cables showed no buckling failure. This indicates that CFRP cable-stayed bridges
have better wind resistance than steel cable-stayed bridges.
Figure 19. (a) Transverse bending moment of girder under wind load; (b) transverse
deformation of girder under wind load.
2.0
(a)
40000
(b)
CFRP
STEEL
20000
Deformation (m)
60000
0
-20000
-40000
-60000
-80000
CFRP
1.5
STEEL
1.0
0.5
0.0
-0.5
300
600
900
1200
1500
x (m)
300
600
x (m)
900
1200
1500
Materials 2014, 7
4873
and the cable response of CFRP cable-stayed bridges under traffic load. The girder deflection and cable
tension response of the bridge were evaluated for different vehicle speeds and levels of road roughness.
Because some of the first natural frequencies of long-span cable-stayed bridges are relatively small, the
calculated results of vehicle-bridge coupling using sophisticated vehicle models show little difference
from those produced using simplified vehicle models. For this reason, the four-degree vehicle model
shown in Figure 20a was used to facilitate calculation. Figure 20b shows the power spectral density of
a sample wave that represents the relatively smooth road surface [41].
Figure 20. (a) Four-degree vehicle model; (b) power spectral density of sample wave.
4
10-4
(a)
m1 (kN-sec /m)
m2 (kN-sec /m)
S()
10-6
1
m1
J1
z1
Input Power spectral density of Roughness
10-8
k4
m5
z5
10-10
0.0
0.1
1.0
k1
c4
c1
m2
z2
k5
c5
k2
c2
Road Surface
10.0
(c/m)
F2
2
m5 (kN-sec /m)
2
2
J 1 (kN-sec -m )
k1 (kN/m)
k 2 (kN/m)
k 4 (kN/m)
k 5 (kN/m)
c1 (kN-sec/m)
c2 (kN-sec/m)
c4 (kN-sec/m)
c5 (kN-sec/m)
1 (m)
2 (m)
17.0
0.70
2.30
40.82
1200.0
1500.0
12000.0
8800.0
17.0
2.0
140.0
7.0
2.84
1.01
F1
Figure 21 gives the flexural deflection responses of the bridge girder under moving traffic at speeds
of 60 km/m, 80 km/m and 100 km/h. The horizontal axis represents the location of the vehicle,
and the vertical axis represents the girder deflection. For both a CFRP cable-stayed bridge and a steel
cable-stayed bridge, the deflection responses at mid-span are dominated by static deflection and a
long-period vibration response, and the short-period vibration response is insignificant. The influence of
traffic speeds was found to be very limited. Figure 21 shows that the flexural deflection of CFRP
cable-stayed bridge is greater than that of the steel-cable-stayed bridge. This is because the CFRP
cable-stayed bridge has relatively less structural rigidity, and the deflection response is mainly caused
by static deformation.
Figure 21. (a) Deflection response of CFRP cable-stayed bridge; (b) deflection response of
steel-cable-stayed bridge.
-10
-10
(a)
(b)
0
60km/h
10
Dy (mm)
Dy (mm)
80km/h
20
100km/h
30
60km/h
10
80km/h
20
100km/h
30
40
40
500
1000
1500
x (m)
2000
2500
3000
500
1000
1500
x (m)
2000
2500
3000
Figure 22 shows the static and dynamic tension responses of cable C1 of a cable-stayed bridge with
moving traffic. The tension response of the CFRP cable was generally similar to that of the steel cable,
even though the natural frequencies of CFRP cable and steel cable are different. This indicates that the
longitudinal vibration of the cable is mainly caused by the vibrations of the support structure, i.e.,
Materials 2014, 7
4874
the anchorage points at the girder and the pylon. The local vibration of the cable does not influence
this response.
Figure 22. Static and dynamic tension responses of cables: (a) CFRP cable under vehicles
moving at 60 km/h; (b) steel cable under vehicles moving at 60 km/h; (c) CFRP cable under
vehicles moving at 80 km/h; (d) steel cable under vehicles moving at 80 km/h; (e) CFRP
cable under vehicles moving at 100 km/h; (f) steel cable under vehicles moving at 100 km/h.
200
0
0
1000
2000
200
3000
x (m)
-200
400
1000
dynamic
2000
dynamic
static
400
static
200
0
0
1000
2000
3000
T (kN)
T (kN)
200
-200
0
0
1000
2000
3000
-200
x (m)
-400
x (m)
-400
dynamic
400
dynamic
400
static
static
200
T (kN)
T (kN)
3000
x (m)
-200
dynamic
static
400
T (kN)
T (kN)
dynamic
static
400
0
0
-200
1000
2000
x (m)
200
0
0
3000
-200
1000
2000
3000
x (m)
6. Conclusions
To explore the feasibility of using CFRP cables in long-span cable-stayed bridges, this paper
establishes a finite element model of a cable-stayed bridge with an ultimate span of 1400 m. The static
and dynamic characteristics of cable-stayed bridges with CFRP cables are studied here. By comparing
CFRP cables with steel cables in terms of mechanical behavior, some conclusions are drawn:
(1) For the same span cables in a 1,400 m span cable-stayed bridge, the self-weight of CFRP cable
was only about 1/7 that of steel cable. Additionally, the Irvine parameters of CFRP cables are only
about 1/6 that of steel cables. Therefore, the CFRP cable performance in tension is better than that
of steel cables.
(2) For the same cable span, the natural frequencies of CFRP cables are about two times those of
steel cables. Additionally, the optimal viscous damping coefficients of CFRP cables are about
half those of steel cables. Therefore, the CFRP cables can help prevent the cable-bridge
coupling vibration caused in lower modes of the cables. The attenuation of the parametric and
wind fluctuating vibrations is better for CFRP as compared to the steel cables as suggested by
the response amplitudes of the cables. The vibration amplitudes of CFRP cables are less than
half those of steel cables.
Materials 2014, 7
4875
(3) CFRP cable-stayed bridge has relatively less structural rigidity, so the maximum girder deflection
of CFRP cable-stayed bridge is about 1.22-times the steel-cable-stayed bridge. However, the
maximum compressive stress in the girders of CFRP-cable-stayed bridges is 9% lower than
that of steel cable-stayed bridges. Additionally, the minimum stability factor of CFRP
cable-stayed bridges is 5% larger than that of steel cable-stayed bridges.
(4) The CFRP cables have a lesser windward area than steel cables. Therefore, the transverse
bending moment and out-of-plane deformation of mid-span girders is slightly larger for steel
cable-stayed bridges than that for CFRP cable-stayed bridges.
(5) An analysis of vehicle-bridge coupling vibrations showed that replacing steel cables with
CFRP cables has little impact on the vibration response of a cable-stayed bridge.
Using CFRP cables on long-span cable-stayed bridges is facilitatory in terms of mechanical
properties. As the manufacturing process matures and the cost decreases, CFRP cables will see more
use in long-span cable-stayed bridges.
Acknowledgments
We gratefully acknowledge the support from the National Natural Science Foundation of China
(No. 50578141), without whom the research would not have taken the present shape.
Author Contributions
In this paper, Xu Xie finished the static calculation of CFRP cables and the cable-stayed bridge, and
wrote the abstract, introduction, conclusions, structural description of the cable-stayed bridge, static
and dynamic characteristics of cable-stayed bridge with CFRP cables. Yonggang Shen finished the
dynamic calculation of CFRP cables and the cable-stayed bridge, and wrote the dynamic
characteristics of CFRP cables and cable-stayed bridge. Xiaozhang Li finished the damping test of
cables, and wrote mechanical properties of CFRP cables.
Conflicts of Interest
The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
Lan, C.M. Study on Life-cycle Safety Assessment Methods for Parallel Wire Stay Cable.
Ph.D. Thesis, Harbin Institute of Technology, Harbin, China, February 2009.
Meier, U. Carbon Fiber-Reinforced Polymers: Modern Materials in Bridge Engineering.
Struct. Eng. Int. 1992, 2, 712.
Meier, U.; Meier, M. CFRP Finds Use in Cable Support for Bridges. Modern Plast. Int. 1996, 4, 8788.
Meier, U. Structural Tensile Elements Made of Advanced Composite Materials. Struct. Eng. Int.
1999, 4, 281285.
Konno, S.; Noro, T.; Yamazaki, S. Material Properties of Carbon Fiber Cables for Cable
Supported Bridges. Bridge Found. Eng. 1999, 9, 2932. (In Japanese)
Materials 2014, 7
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
4876
Take, N.; Higashino, T.; Mastsumura, T.; Nishimura, N. Study on Aerodynamic Stability and
Preliminary Design of Dual Cable Suspension Bridges Using Advanced Composites. Trans. Jpn.
Soc. Comput. Eng. Sci. 1999, 1, 8994. (In Japanese)
Mei, K.; Lu, Z. Application Prospect of CFRP to Super Length Suspension Bridge and
Cable-Stayed Bridge. Bridge Constr. 2002, 2, 7578. (In Chinese)
Kou, C.; Xie, X.; Kao, C.; Huang, J.Y. Static Behavior of Long-span Cable-stayed Bridges Using
Carbon Fiber Composite Cable. J. Zhejiang Univ. Eng. Sci. 2005, 39, 137142. (in Chinese).
Cheng, S.H.; Lau, D.T. Impact of Using CFRP Cables on the Dynamic Behavior of Cable-stayed
Bridges. Reported at the IABSE Symposium, Budapest, Hungary, 1115 September 2006.
Wu, Z.S.; Wang, X. Investigation on a thousand meter scale cable-stayed bridge with fiber
composite cables. In Proceedings of the 4th International Conference on FRP Composites in Civil
Engineering, Zurich, Switzerland, 2224 July 2008.
Wen, X.; Cai, C.S.; Xiao, R.C.; Lu, D. Study of Super Long Span Cable-stayed Bridges with CFRP
Components. Struct. Eng. 2011, 33, 330343.
Li, H.; Huang, Y.; Chen, W.L.; Ma, M.L.; Tao, D.W.; Ou, J.P. Estimation and Warning of Fatigue
Damage of FRP Stay Cables Based on Acoustic Emission Techniques and Fractal Theory. Comput.
Aided Civil Infrastruct. Eng. 2011, 26, 500512.
Wen, X.; Cai, C.S.; Xiao, R.C.; Lu, D. Concept and Analysis of Stay Cables with a CFRP and
Steel Composite Section. Struct. Eng. 2012, 16, 107117.
Meier, U. Carbon Fiber Reinforced Polymer Cables: Why? Why Not? What If? Arabian J. Sci. Eng.
2012, 37, 399411.
Seible, F.; Karbhari, V.M.; Burgueno, R. Kings Stormwater Channel and I-5/Gilman Bridge, USA.
Struct. Eng. Int. 1999, 9, 250253.
Karbhari, V.M.; Seible, F. Fiber-Reinforced Polymer Composites for Civil Infrastructure in the
USA. Struct. Eng. Int. 1999, 4, 274277.
Burgoyne, C.J. Advanced Composites in Civil Engineering in Europe. Struct. Eng. Int. 1999, 4,
267273.
Christoffersen, J.; Hauge, L.; Bjerrum, J. Footbridge with Carbon-fibre-reinforced Polymers,
Denmark. Struct. Eng. Int. 1999, 9, 254256.
Lu, Z.T.; Mei, K.H. First Application of CFRP Cables for a Cable-stayed Bridge in China. China
Civil Eng. J. 2007, 40, 5059. (In Chinese)
Yamgauchi, H.; Ito, M. Mode-dependence of Structural Damping in Cable-stayed Bridges.
J. Wind Eng. Ind. Aerodyn. 1997, 72, 289300.
Zhang, H.; Xie, X.; Zhao, J. Parametric Vibration of Carbon Fiber Reinforced Plastic Cables with
Damping Effects in Long-span Cable-stayed Bridges. J. Vib. Control 2011, 17, 21172130.
Xie, X.; Nakamura, H.; Maeda, K.; Zhang, Z.; Enomoto, T. Theoretical Analysis and
Experimental Test on Damping Characteristics of CFRP Stay Cables. Eng. Mech. 2010, 27,
205216. (In Chinese)
Kady, H.M.; Arockiasamy, M.; Samaan, S. Damping Characteristics of Carbon Fiber Composite
Cables for Application in Cable-stayed Bridges. In Proceedings of IABSE Conference on
Cable-stayed Bridges- Past, Present and Future, Malmo, Sweden, 24 June 1999.
Materials 2014, 7
4877
24. Irvine, H.M. Cable Structures; MIT press: Cambridge, MA, USA, 1981.
25. Caetano, E.d.S. Cable Vibrations in Cable-Stayed Bridges; Structural engineering documents 9.
IABSE-AIPC-IVBH: Zurich, Switzerland, 2007.
26. Takahashi, K. Dynamic Stability of Cables Subjected to An Axial Periodic Load. J. Sound Vib.
1991, 144, 323330.
27. Uhrig, R. On Kinetic Response of Cables of Cable-stayed Bridges due to Combined Parametric
and Forced Excitation. J. Sound Vib. 1993, 165, 182192.
28. Lilien, J.L.; Da Costa, A.P. Vibration Amplitudes Caused by Parametric Excitation of Cable
Stayed Structures. J. Sound Vib. 1994, 174, 6990.
29. Kang, Z.; Zhong, W.X. Numerical Study on Parametric Resonance of Cable in Cable Stayed
Bridge. China Civil Eng. J. 1998, 31, 1422. (In Chinese)
30. Sun, B.N.; Wang, Z.G.; Ko, J.M.; Ni, Y.Q. Parametrically Excited Oscillation of Stay Cable and
Its Control in Cable-stayed Bridges. J. Zhejiang Univ. Sci. 2003, 4, 1320.
31. Yang, S.Z.; Chen, A.R. Parametric Oscillation of Super Long Stay Cables. J. Tongji Univ. Nat. Sci.
2005, 33, 13031307. (In Chinese)
32. Georgakis, C.T.; Taylor, C.A. Nonlinear Dynamics of Cable Stays. Part 1: Sinusoidal Cable
Support Excitation. J. Sound Vib. 2005, 281, 537564.
33. Georgakis, C.T.; Taylor, C.A. Nonlinear Dynamics of Cable Stays. Part 2: Stochastic Cable
Support Excitation. J. Sound Vib. 2005, 281, 565591.
34. EI Ouni, M.H.; Kahla, N.B.; Preumont, A. Numerical and Experimental Dynamic Analysis and
Control of a Cable Stayed Bridge under Parametric Excitation. Eng. Struct. 2012, 45, 244256.
35. Takahashi, K.; Konishi, Y. Non-linear Vibrations of Cables in Three Dimensions, Part I:
Non-linear Free Vibrations. J. Sound Vib. 1987, 118, 6984.
36. Takahashi, K.; Konishi, Y. Non-linear Vibrations of Cables in Three Dimensions, Part II:
Out-of-plane Vibrations under Inplane Sinusoidally Time-varying Load. J. Sound Vib. 1987, 118,
8597.
37. Esaka, Y.; Matsui, T.; Qiao, F.; Moribe, Y.; Sugiyama, E. A Study on Response of Seismically
Isolated Large Span Domes to Fluctuating Wind Loads. Part 1: Wind Response of Base Isolation
System; Architectural Institute of Japan: Tokyo, Japan, 2001. (In Japanese)
38. Xie, X.; Zhang, Z.C.; Nakamura, H.; Maeda, K. Nonlinear vibration of stay cables under
fluctuating wind load and support excitation. China Civil Eng. J. 2008, 41, 6672. (In Chinese)
39. Xie, X.; Zhu, Y.F. Study on Behavior of Long-span CFRP Cable-stayed Bridges. Eng. Mech. 2007,
24, 113120. (In Chinese)
40. Nagai, M.; Asano, K.; Watanabe, K. Applicability of the Ef method and design method for
evaluating the load-carrying capacity of girders in cable-stayed bridges. J. Struct. Eng. JSCE 1995,
41A, 221228. (In Japanese)
41. Xie, X.; Zhu, Y.F.; Shen, Y.G. Study on vibration of long-span cable-stayed bridge with steel and
CFRP cables due to moving vehicles. Eng. Mech. 2007, 24, 5361. (In Chinese)
2014 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).